Without warning, X has started training its Grok AI-powered chat platform using publicly posted content from its users, processing their data without initial notification or consent.
As artificial intelligence platforms vie for supremacy, their massive language models are perpetually seeking out new data to refine and improve their performance.
Having your information makes my insights incredibly valuable. Regardless of consent, many online platforms exploit users’ personal data without their knowledge or explicit permission from the original sources.
To avoid being overlooked in the sport, X surreptitiously began coaching its Grok AI chat platform by incorporating users’ posts without seeking permission or issuing a public statement regarding the change.
It wasn’t until Thursday, July 25, that Facebook introduced a novel feature under its privacy settings, allowing the platform to utilize user data with their explicit consent. What’s even more concerning is that this setting is enabled by default, rather than requiring explicit user consent to turn it on.
X’s security team announced in a tweet that the new feature is initially available only on the website version, with a rollout to mobile devices underway.
“All X customers have the flexibility to control whether their public posts are used to train Grok, our AI-powered search assistant.”
“This consent grants permission for your interactions, inputs, and outcomes related to Grok to be used, in addition to the existing controls you have established.” Accessible online, this setting can be rapidly deployed across networks.
Accessible to all users, this privacy option can be easily found under the same Privacy settings used by every individual, requiring merely a few simple clicks to navigate.
The website’s webpage has also been updated to note that X utilizes your submissions to educate Grok, a feature previously absent from the original content.
“To maximize your expertise, leverage our platform’s curated content – including your X posts, person-to-person interactions, and tangible outcomes – as a catalyst for continuous learning and refinement through personalized coaching and fine-tuning.”
Since this setting is enabled by default, there’s a strong possibility that X has been training its AI platform using member data for quite some time now. Regardless, this new opportunity grants customers the flexibility to make informed choices.
Grok’s potential for growth and self-actualization lies in his ability to assimilate and adapt to new situations. Therefore, deciding what knowledge to share with him necessitates a consideration of his developmental readiness.
Do not simply impart arbitrary facts or skills without first considering their relevance and impact on Grok’s overall well-being and cognitive maturation?
If you’re unwilling to allow X to utilize your content for coaching Grok, you have a couple of options:
While making your account personal can be an option, it may inherently restrict the breadth of your knowledge on X. It’s not a viable alternative.
To increase the chances of successful login, try accessing the X website from your desktop and clicking on the icon located at the bottom-right corner of your screen.
To achieve this, navigate to the relevant section by clicking on then selecting beneath the categories of Knowledge sharing and personalization, and subsequently disable the “setting” option, as demonstrated below.
Once this setting is activated, the utilization of your data for training Grok and sharing it with Musk’s AI company, xAI, will cease immediately.
